Consider this output from nosetests:
$ nosetests
F.......
======================================================================
FAIL: test (failing_test.ClassWithTest)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/mambocab/py/nose-pasteable/test/failing_test.py", line 10, in test
assert False
AssertionError
----------------------------------------------------------------------
Ran 8 tests in 0.009s
FAILED (failures=1)How do you go about running just that failed test? Well, you copy and paste the test's name into your command prompt...
$ nosetests test (failing_test.ClassWithTest)
zsh: invalid mode specification
$ # oh come onNo. No you don't. The way nose outputs tests names looks pretty good, but is not useful. This has bugged me ever since @tenderlove pointed out his favorite feature of rspec: you can just copy and paste its output into the command line to rerun failing tests.
So, with nose-pasteable, the output you get is a little nicer. Enable it with the --pasteable option, or by setting NOSE_PASTEABLE to a non-empty string in the environment in which you run nose:
$ nosetests --pasteable
F.......
======================================================================
FAIL: test/failing_test.py:ClassWithTest.test
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/mambocab/py/nose-pasteable/test/failing_test.py", line 10, in test
assert False
AssertionError
----------------------------------------------------------------------
Ran 8 tests in 0.009s
FAILED (failures=1)
$ # copy and paste that thing from up there...
$ nosetests test/failing_test.py:ClassWithTest.test
F
======================================================================
FAIL: test (failing_test.ClassWithTest)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/mambocab/py/nose-pasteable/test/failing_test.py", line 10, in test
assert False
AssertionError
----------------------------------------------------------------------
Ran 1 test in 0.000s
FAILED (failures=1)Heck yes. One little annoyance gone.
Installation
Easy: use pip install nose-pasteable. Slightly harder: clone this repo, then use pip install . from inside it.
Development
Clone this repository, then create a branch in a repository you control. GitHub is most convenient for me, but I don't judge. GitLab is awesome, and BitBucket is pretty cool too. Go for it! I accept patches!
Do your development on that branch, adding tests in test_script.sh as you go. To run the tests, install bats by following the instructions there. Then, from the top-level directory of this repository, run ./test_script.sh. All existing tests should pass, and any new features should at least have a smoke test and a test that makes sure they don't run when they're disabled.
Then, let me know about the changes, either through a pull request or a link in an issue.
